@yura

people keep asking why I chose the @quotient score over @neynar, so here’s the short answer before launch, I ran a quick sanity check to see which score aligned better with how the market actually prices attention on @surgelabs. I took the top 1,000 sellers on ampsdotfun (hello @phil) and compared their historical average price_per_reaction with two scores (Neynar and Quotient) using Spearman rank correlation. Result: • Neynar vs price_per_amp: 0.098 • Quotient vs price_per_amp: 0.181 Quotient showed a meaningfully stronger rank correlation with pricing, which was enough for me to go with it. This kind of analysis is far from perfect and has plenty of limitations, but it gave me enough signal to make the call. P.S.: this is not a knock on Neynar at all. I LOVE it with a passion. Surge wouldn’t exist without Neynar, and its score is excellent for many other use cases. This choice was purely about fit for this specific problem.

8 replies

10 recasts

60 reactions